On the effects of grid size and shape when mapping the distribution range of a recolonising wolf (Canis lupus) population

Abstract

An estimate of changes in a species' distribution range is a key variable in assessing its conservation status. It may be based on the direct detection of individuals, or on the use of indirect presence sign surveys. In both cases, the process requires one to switch from a point-based approach, where individuals/presence signs are located using a coordinate system, to an area-based one, each original point being replaced by a cell area unit (CAU), with a given shape and size. The estimated distribution range (EDR) is the spatial union of the CAUs over the area of interest. Based on wolf presence signs collected in France (1996-2006), we analysed the influence of the shape and size of types of CAUs (circular area versus square grid mesh; 6, 25, 50 and 100 km2) on the changes in EDR. EDR increased with time and a saturating phase was noticed by the end of the period. We assessed the effects of the year and the type of CAU on EDR using exploratory data analysis. Larger CAUs resulted logically in larger EDR values, whatever the CAU shape. For a given CAU size, contiguous square grids yielded larger EDR values than overlapping circular buffers. The effect of the interactions between the year and the type of CAU on EDR changes was evidenced using an auto-modelling method based on principal component analysis. Compared to smaller units, larger CAUs resulted in larger growth rates during the range increase phase, and in smaller rates during the saturating phase. A basic and descriptive conceptual model helped interpreting this pattern as a consequence of the characteristics of the colonisation process in the wolf population. We discuss the present results within the framework of conservation status assessment and management of the wolf population. © 2010 Springer-Verlag.